NEWVectors or files. Pick a path.Start →
    Back to Videos

    Quantisation Scrambles What Is Near What

    61
    Multimodal Search in Action
    Mixpeek Team
    August 2, 2026

    Summary

    Generative recommendation turns items into discrete tokens so a model can generate the next item the way a language model generates the next word, and making those tokens means quantising each item embedding into a short code. In the pretrained embedding space similar items sit near each other, and that adjacency is the signal; after quantisation the intrinsic structure is disrupted, so the model's perception of item similarity is distorted before training begins. TopoTok (RecSys 2026) measures that distortion and repairs it with multi level distillation, coarse to fine: inter group for global cluster relations, intra group for local structure, inter item for fine grained alignment, reporting up to 9.42% better Recall@5. One distinction matters: this is not index-time vector compression, where you rescore the shortlist at full precision to recover accuracy. Here the structure is gone before the model trains, so there is nothing to rescore. Paper: arxiv.org/abs/2607.18600

    recsyssemantic-idsquantizationembeddingsgenerative-recommendationinformation-retrieval

    About this video

    Generative recommendation turns items into discrete tokens so a model can generate the next item the way a language model generates the next word, and making those tokens means quantising each item embedding into a short code. In the pretrained embedding space similar items sit near each other, and that adjacency is the signal; after quantisation the intrinsic structure is disrupted, so the model's perception of item similarity is distorted before training begins. TopoTok (RecSys 2026) measures that distortion and repairs it with multi level distillation, coarse to fine: inter group for global cluster relations, intra group for local structure, inter item for fine grained alignment, reporting up to 9.42% better Recall@5. One distinction matters: this is not index-time vector compression, where you rescore the shortlist at full precision to recover accuracy. Here the structure is gone before the model trains, so there is nothing to rescore. Paper: arxiv.org/abs/2607.18600

    Frequently Asked Questions